And in fact that should be the goal — maximizing system complexity, minimizing system definition (perhaps there are better terms for this…)
Games that manage to do so successfully have the opportunity to be replayed through multiple lifetimes
I think this is less than half true. People wouldn't care about those games, if they didn't have the amount of emergent complexity that they do. It would be more accurate to say that they soak up so much attention because of legacy culture and competitions. But those wouldn't exist, if those games didn't have certain qualities which made them eminently interesting and replayable in the first place.
Could they have been replaced through different accidents of history with other games entirely? Sure. But those games would have had many of the same qualities.
Fixed start, no hidden information, turn based games are pretty rare anymore. They're not eyecatching, but they're still good concepts.
WARNING: KINDA A RAMBLE, STOP READING HERE IF BORED EASILY.
I've also played a lot of Magic the Gathering, which of course has hidden information and a hybrid fixed start. Both players have the same number of cards in their hand and on the field[1], but will almost certainly have different cards to each other even in Deck X vs Deck X mirror match ups.
In competitive MtG with a best-of-3 system, there can be a sense felt by the players that the die roll for Game-1-Turn-1 has a disproportionate effect on the match. Despite the roll giving you the choice of first or second turns, it is decidedly unusual for players to win the roll and choose to go second (even though the player who goes first skips their first card draw).
For the sake of the argument, if we assume going second definitively has lower chance of victory then the players perceptions can become frustrated and this can effect their desire to replay the game.
Because the loser of Game 1 selects who goes first in Game 2, and will likely pick themselves, and the winner of Game 1 is more likely to have won the Game-1-Turn-1 dice roll.
We get an outcome in which beating your opponent in Game 1 cascades, because your opponent will choose to go first in the 2nd game and either win returning the turn 1 advantage to you for the tiebreaker G3 or lose and the match concludes. Because Game 1 has disproportionate significance AND is the only game where luck decides who goes first AND is chronologically the the first to occur, I have observed (and felt) a kind of stewing frustration at 'playing out' the rest of some matches after losing the die roll and then having a poor early game 1.
I believe a this is mostly concerns players emotions and perceptions about 'inherent' advantages, which is why I've talked about it as a factor of replayability rather than just strategy.
[1]In a typical circumstance, some cards or deck types can alter the initial game state of the board or hand.
I also like bidding systems. Bid some in game resource for the right to go first. On ties, pick at random. If the game is popular the meta will converge on a fair bid to null out the advantage, and may evolve over time. Twilight Struggle is a great 1v1 board game that settled on this.
For me it's time. If I know in advance the game will take 10 hours to finish then no matter how much variety it has, I'm not gonna play it again.
The second time, you mostly know what you’re in for. There may be different paths to take, there may be new stuff unlocked by your first play through. But if you made it to the end then you’ve probably exhausted most of its unique content. There may be other things to come back for (DLC, competitive play, 100%, achievements, etc) but it’s easy to put it down for good.
How many books, movies, or TV shows do you come back to? How many do you go through once and never again? Why should a game be any different?
You can design a game such that it's a somewhat unique experience every time you play it. Even games with limited variation in setup allow different approaches from the player. That's not possible with traditional novels and film. At best, maybe you can review a film or a book and pay more attention to some detail you didn't notice that much in the first view, but in the end you are rather passively consuming a fixed narrative.
Of course, how meaningful this is depends on whether the game mechanics stand out on their own, or if it's mostly a vessel for a fixed narrative. If the game is just a mechanically bland walk from one cutscene to another, with little agency, it might not seem worth it. A lot of AAA titles are like this IMO. If the game generates a unique set of challenges every time you play it, or rewards new approaches, there is much more fun to be had replaying it.
This is what I mean by the systems are what matter, not content variation. You can play sim city repeatedly, despite there never being new content, because learning the game systems (inputs, and predicted outputs, and then achieving goals based on that) is what’s valuable, and there’s a lot there to unpack.
Assassin’s creed on the other hand should only be played once, because the game systems are trivial — you’ve probably understood everything it has to offer long before you’ve completed the major storyline. So the only driver is content, and really only the shallow surface novelty of that content (story, aesthetics, etc).
So once you’ve seen it… why see it again?
Same for open-ended building games like Minecraft (technically completed when you kill the dragon), Cities: Skylines, or Factorio (when you send a rocket).
I would rather try a janky, kind of bad, yet mechanically/systems-wise interesting game like Kenshi over another AAA movie-game any day of the week. I have no problem with linear/linearish narrative-focused games—I'm currently playing through a visual novel actually. but interactively conveying a pre-defined narrative (possibly with some "choices" sprinkled in here and there) is only the tiniest tip of the iceberg of possibilities the medium affords!
the problem here is, video games are still a very young medium, so many people have preconceptions about conventional storytelling/narrative structure that they carry over from film and literature. many people want a video game they're considering purchasing to be something with an Hour Count you can look up online, weigh against your interest in the plot/setting/characters/whatever, and decide if playing through the predefined arc of the game is something worth their time. few people look for games with mechanics and systems that converge in interesting, complex, emergent ways, and say to themselves, that looks like something fun to play (with) for as long as I keep enjoying it!
Sea of Thieves is an interesting game because it attempts to bridge this gap. when my buddies and I play it every now and again, we enjoy planning some loosely-defined goals for a session, then just fucking around on the open seas as we see fit. for us, something like seeing how many powder kegs we can gather and fit into the crow's nest so as to test the viability of hucking them into the water and shooting them when fighting player ships, is infinitely more entertaining and interesting that playing through the scripted storyline quests. but on the other hand, my fiancée loves to play through these scripted storyline quests with the wife of one of my friends, just dutifully following the quest instructions, looking up how to do things online if needed, whatever is necessary to check that quest off their list. this same friend's wife has expressed in the past that she cannot for the life of her understand the appeal of the roguelike/lite genre—"why would you just want to play the same thing over and over?"
many people like (and expect games to be) linear narrative experience video games, but more games should try to be more gamey, more systemsy, more emergent, because that's what makes the medium unique and different than others. the only issue with this is player expectations, which hopefully will shift somewhat with time.
This usually occurs because the game systems themselves are trivial (often a set of independent features stuffed behind the same UI, but A has no influence on B), because one or a few of the systems are overwhelmingly influential (e.g. melee mechanics don’t matter if guns are trivially superior for all scenarios), or because the game itself doesn’t incentivize the player to actually exploit those systems to any real degree (e.g. the game is too easy or the levels don’t introduce enough new constraints, so the first solution found applies globally; often fixed by player-induced artificial challenges like no X runs).
Multiplayer however cheats this because even trivial systems can become a huge exploration space when incentivized by direct competition (analysis and understanding is pushed to the limit; e.g. sm64 requires a certain level of understanding to play effectively, but speed runners and TAS take that understanding to the nth degree) and the players themselves bring metagaming as an additional system worth exploring (eg any competitive multiplayer)
It’s also confounded by multiplayer games that are less “played” and more used as a social space — replacing the shopping mall — like many mmo’s, Minecraft, etc. You’re not replaying so much as visiting and killing time between social interactions (so even highly repetitive actions often suffice).
The content variation is largely necessary to give reason and space to explore the game systems under novel constraints.